In the 1890s, public transport systems in the United States were often streetcars - trams. These, of course, were electrically powered rail vehicles drawing current from overhead wires. A very popular system of collection was a long pole with a small grooved wheel at the wire end: this was known as a trolley and, by extension, the vehicle and the system also became known as a trolley.

To generate traffic at weekends (when people were travelling less for work), the transit companies would create attractions - in particular, recreation areas at the ends of their lines where land was cheap. Over 100 of these Trolley Parks were opened.

Someone returning from that era would have a shock if they found a trolley park today - a place where self-help trolleys are stored in supermarkets or at airports!

How times change - and language with it!

The only thing you can be almost certain about with a forecast is that it'll be wrong.

Plans for infrastructure - particularly public transport infrastructure - have often been criticised for over-optimism.

There is a book - "Megaprojects and risk" - about some notable failures to get the forecasts reasonably right. In my view it damaged its credibility in particular by citing Channel Tunnel experience - that project was just so political that no forecasts were likely to be reliable. It also looked at the Oresund Fixed Crossing and added to the outturn costs the Malmo City Tunnel, a free-standing project (not yet completed) which is not part of the Oresund project at all.

Highway infrastructure projects have also been criticised recently for bad forecasting. In "Infrastructure Investor"'s review of the year 2009, Michael Dinham of ING Bank notes that only two of the 30 toll roads the bank has lent money to have met their original forecasts, and under-performance on the rest has averaged 40%. The range is 10% to 80%. Roads, he concludes, are the worst-performing sub-sector within infrastructure.